In testing of programmable controllers, manual selection of test cases is still the most common method in practice. This is however tailor-made, time consuming and error-prone. Traditional model-based methods can hardly handle industrial scale systems which usually possess a significant number of states, and signals of sensors and actuators. In this paper, we propose a model-based testing framework that utilizes simplified plant features to reduce the number of test cases, and at the same time also guarantees a full coverage of nominal behavior of system under test. The proposed framework has been illustrated on a case study.     «

In testing of programmable controllers, manual selection of test cases is still the most common method in practice. This is however tailor-made, time consuming and error-prone. Traditional model-based methods can hardly handle industrial scale systems which usually possess a significant number of states, and signals of sensors and actuators. In this paper, we propose a model-based testing framework that utilizes simplified plant features to reduce the number of test cases, and at the same tim...     »